Journal of Affective Disorders 148 (2013) 435–439
Contents lists available at SciVerse ScienceDirect
Journal of Affective Disorders
journal homepage: www.elsevier.com/locate/jad
Brief Report
Risk factors associated with repetition of self-harm in black and
minority ethnic (BME) groups: A multi-centre cohort study
Jayne Cooper a,n, Sarah Steeg a, Roger Webb a, Suzanne L.K. Stewart d, Eve Applegate a,
Keith Hawton b, Helen Bergen b, Keith Waters c, Navneet Kapur a
a
Centre for Mental Health and Risk, University of Manchester, Jean McFarlane Building, Oxford Road, Manchester M13 9PL, UK
Centre for Suicide Research, University of Oxford, Department of Psychiatry, Warneford Hospital, Oxford OX3 7JX, UK
c
Mental Health Liaison Team, Rehabilitation Centre, Royal Derby Hospital, Derbyshire Healthcare NHS Foundation Trust, Derby, UK
d
Department of Psychology, University of Chester, Parkgate Road, Chester, CH1 4BJ, UK
b
a r t i c l e i n f o
abstract
Article history:
Received 6 January 2012
Received in revised form
19 September 2012
Accepted 6 November 2012
Available online 8 December 2012
Background: Little information is available to inform clinical assessments on risk of self-harm repetition
in ethnic minority groups.
Methods: In a prospective cohort study, using data collected from six hospitals in England for self-harm
presentations occurring between 2000 and 2007, we investigated risk factors for repeat self-harm in
South Asian and Black people in comparison to Whites.
Results: During the study period, 751 South Asian, 468 Black and 15,705 White people presented with
self-harm in the study centres. Repeat self-harm occurred in 4379 individuals, which included 229
suicides (with eight of these fatalities being in the ethnic minority groups). The risk ratios for repetition
in the South Asian and Black groups compared to the White group were 0.6, 95% CI 0.5–0.7 and 0.7, 95%
CI 0.5–0.8, respectively. Risk factors for repetition were similar across all three groups, although excess
risk versus Whites was seen in Black people presenting with mental health symptoms, and South Asian
people reporting alcohol use and not having a partner. Additional modelling of repeat self-harm count
data showed that alcohol misuse was especially strongly linked with multiple repetitions in both BME
groups.
Limitations: Ethnicity was not recorded in a third of cases which may introduce selection bias.
Differences may exist due to cultural diversity within the broad ethnic groups.
Conclusion: Known social and psychological features that infer risk were present in South Asian and
Black people who repeated self-harm. Clinical assessment in these ethnic groups should ensure
recognition and treatment of mental illness and alcohol misuse.
& 2012 Elsevier B.V. All rights reserved.
Keywords:
Self-harm
Ethnic minority
Repetition
Risk factors
1. Introduction
Rates and risk factors for self-harm and suicide vary amongst
Black and Minority ethnic (BME) groups within the UK compared
to White groups, including between different age and sex groups
(Bhui et al., 2007; Cooper et al., 2010). Higher rates of self-harm
have previously been reported in South Asian females compared
to South Asian males or White females (Cooper et al., 2006; Bhui
et al., 2007). In a more recent study based on the Multicentre
Study of Self-harm in England we found that rates of self-harm
were highest in young Black females (pooled rate ratio for Black
females aged 16–34 years compared with White females 1.70,
95% CI 1.5–2.0) (Cooper et al., 2010). BME groups experience
socioeconomic inequalities which have been linked to subsequent
n
Corresponding author. Tel.: þ44 161 275 0718; fax: þ 44 161 2750716.
E-mail address: [email protected] (J. Cooper).
0165-0327/$ - see front matter & 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jad.2012.11.018
inequalities in health (Nazroo et al., 2007), and racial/ethnic
discrimination has a strong association with common mental
disorders. (Bhugra and Arya, 2005). Little is known about risk
factors for repetition of self-harm in BME groups that can be used
to facilitate appropriate clinical management and suicide prevention measures (Kapur et al., 2006). We aimed to identify risk
factors for repeat self-harm in Black and South Asian people, the
largest ethnic minority groups in the UK, following an index selfharm presentation to hospital emergency departments (EDs) in
three centres in England.
2. Methods
A prospective cohort study was carried out using data collected from six hospitals in Manchester (3), Oxford (1) and Derby
(2) for self-harm presentations occurring between 2000 and 2007.
An established monitoring system in each centre was used to
436
J. Cooper et al. / Journal of Affective Disorders 148 (2013) 435–439
retrieve data, as described in detail previously (Cooper et al.,
2010; Bergen et al., 2010a). Self-harm attendances were identified
via detailed examination of computerised emergency department
records and defined consistently across all three centres as
intentional self-poisoning or self-injury, irrespective of motivation and degree of suicidal intent. Most participants received a
psychosocial assessment, and clinicians recorded a wide range of
sociodemographic and clinical information using research assessment forms. For participants who were not assessed, basic
information was collected by research clerks from medical
records. Ethnicity was ascribed by the treating clinician at time
of admission according to standard UK national 2001 census
categories, or later attributed using information from ED patient
record systems. For analysis purposes, we were interested in
three broad groups: Black people, (Black African-Caribbean and
Black Other), South Asian people (Indian, Pakistani and Bangladeshi origin) and White people as a reference group. A lower age
limit of 16 years was applied as service provision for under
16-year-olds differs from that in over 16-year-olds. We used a
comprehensive data-linkage process (Murphy et al., 2012) to
identify repeat self-harm and subsequent suicides. The Medical
Research Information Service identified deaths that occurred in
the UK by suicide (ICD-10 codes X60 to X84) and undetermined
cause (ICD-10 codes Y10 to Y34 excluding Y33.9) up to December
2009. These were included in the ‘repeater’ group.
2.1. Ethical considerations
Oxford and Derby both have approval from local National Health
Service (NHS) research ethics committees to collect data on selfharm for local monitoring and multicentre projects. The monitoring
of self-harm in Manchester is part of a clinical audit system and has
been ratified as such by local research ethics committees. Thus,
formal ethics committee approval was not required for that centre.
All centres are fully compliant with the UK Data Protection Act 1998,
and have support under section 251 of the NHS Act 2006 regarding
the use of patient-identifiable information.
3. Analysis
All analyses were performed using STATA v10 (Statacorp,
2007). We calculated risk ratios for first episode repetition in
the South Asian and Black groups compared to Whites using logbinomial regression models, adjusted for age, sex, method of
harm and drugs used in self-poisoning (paracetamol/antidepressants/benzodiazepines) as potential confounders. We used Cox
proportional hazard models to determine hazard ratios and their
confidence intervals for repetition, for Black, South Asian and
White individuals combined. Before taking any additional confounders into consideration, we fitted a 3-category ‘centre’ variable as a potential confounder in all of our regression models to
account, as far as possible, for between-centre differences. Interaction terms were fitted to examine heterogeneity of factors
associated with risk in the Black and South Asian groups verses
Whites, and the significance of these terms were formally tested
using Wald tests. Stratified analyses were carried out to obtain
ethnic group-specific hazard ratios. Aggregated data were used in
Poisson regression models to generate incidence rate ratios (IRRs)
to examine variables associated with multiple episodes of repeated
self-harm. A substantial proportion of these models showed overdispersion (using the alpha¼0 test as assessed in equivalent
negative binomial regression models) (Gardner et al., 1995), so
we generated IRRs based on negative binomial regression. We
included up to ten repeat episodes within the study period for each
individual, as in our previous research (Steeg et al., 2012).
4. Results
Between 2000 and 2007, ethnicity was known in 17,324/
25,328 (68.4%) individuals presenting to the participating EDs.
Of these, 1619 (9.3%) were identified as being from an ethnic
minority background: 751 from South Asian and 468 from a Black
ethnic group. The remaining 400 individuals were from other
ethnic groups and were not included because of small numbers in
individual subgroups (including mixed, Chinese and other Asian
origin). There were 15,705 Whites. Of the 4379 individuals in the
study cohort who repeated self-harm, there were 8 suicides in the
BME groups (Black: 2 males, 1 female; South Asian: 3 males,
2 females), and 213 suicides in the White group of which 65%
were males. The risk ratios for repetition in the South Asian and
Black groups compared to the White group separately, adjusting
for baseline characteristics, were 0.6, 95% CI 0.5–0.7 and 0.7, 95%
CI 0.5–0.8, respectively. The mean number of days to the first
repeat ED presentation following an index self-harm episode was
293 for South Asian people (interquartile range 16–459; median¼92) and 366 for Black people (interquartile range 43–551;
median 173) and 339 days for White people (interquartile range
30–476; median 152). Due to the skewed distributions we
formally compared these median values: there was no significant
difference between South Asian and White people (p¼0.21) or
between Black and White people (p ¼0.73).
Six factors were identified as having a significant effect in South
Asians versus Whites (Table 1): not having a partner, problems with
alcohol as a precipitant to the act, use of alcohol at the time of selfharm, self-harm in the past year, receiving psychiatric treatment at
the time of the act and using cutting or stabbing as the method of
self-harm. These variables all showed larger effect sizes in the South
Asian group. There was one factor that emerged as having a different
effect in the Black group versus Whites: problems with mental
health as a direct precipitant to the self-harm which were more
common in Black individuals who repeated. A sensitivity analysis
was conducted to see if the results were altered by including the
category ‘other’ ethnic group in the reference group, with the results
remaining essentially unchanged. When we examined heterogeneity using negative binomial regression to take multiple repetitions
into account, we found that problem use of alcohol as a precipitant
to self-harm and alcohol use at the time of the act showed a much
stronger effect in both BME groups than in the Cox proportional
hazard models. Additionally, in South Asians, not having a partner
and having self-harmed in the past year were found to be more
strongly linked with multiple repetition than was the case for
Whites Table 2.
5. Discussion
To our knowledge this is the first study to examine risk factors
for repetition of self-harm in South Asian and Black groups in the
UK. The risk of repetition was significantly lower in the ethnic
minority groups compared to the White group. We found that
those who repeated in the South Asian and Black groups had
social and clinical characteristics similar to patients who repeated
in general (Kapur et al., 2006). This is in contrast to the varied
profile of BME groups at their index episode of self-harm (Cooper
et al., 2010). However there was some differentiation of risk
factors for repetition between the groups. Increased risk of
repetition compared to the White group was found in South
Asians who did not have a partner, had problems with alcohol as a
precipitant to the act, used alcohol at the time of self-harm, had
previously self-harmed in the past year, were receiving psychiatric treatment at the time of the index episode and used cutting or
stabbing as a method of self-harm. In the Black group, problems
J. Cooper et al. / Journal of Affective Disorders 148 (2013) 435–439
437
Table 1
Cox proportional hazards models showing hazard ratios for time to repetition from index self-harm episode, with significant interactions between risk factor and
ethnic group.
Variableb
White, Black
and SA
Repeaters
(%)
Total
Sociodemographic:
In employment or education
Unemployed
HRa
(95% CI)
4379
(25.9)
3153
(24.7)
1226
(29.4)
1966
1.0
(23.6)
800 (30.1) 1.4 (1.3 to 1.5)
Psychiatric treatment and previous self-harm:
No self-harm in the past year
2410
(21.6)
Self-harm in the past year
1184
(36.3)
No previous self-harm
Self-harm 41 year ago
Not in current treatment
In current treatment
No previous treatment
Any previous treatment
Circumstances of the act:
Self-poisoning involved
Self-cutting/stabbing involved
Alcohol used in the act
P value for
interaction
term:
SA vs. White
Repeaters
(%)
HRa
(95% CI)
P value for
interaction
term:
Black vs. White
79 (16.9)
1.3 (1.2 to 1.4)
Lives with family/friends
674 (29.5)
834 (29.9)
1419
(20.8)
895 (24.0)
HRa
(95% CI)
1.0
807 (22.3) 1.0
2245
1.2 (1.1 to 1.3)
(26.4)
Factors precipitating self-harm:
Problem alcohol use
Mental health symptoms
Relationship problems with
partner
Relationship problems with
family/friend
Repeaters
(%)
Black
107 (14.2)
Married or partnered
No partner
Lives alone or homeless
South Asian
21 (8.6)
63 (19.2)
1.0
2.4 (1.4 to 3.9) 0.006
1.2 (1.1 to 1.3) 10 (33.3)
1.2 (1.1 to 1.3)
0.7 (0.6 to 0.7)
2.9 (1.4 to 5.8) 0.009
18 (28.6)
2.1 (1.2 to 3.6) 0.03
0.9 (0.8 to 1.0)
1.0
53 (10.2)
1.9 (1.8 to 2.0) 30 (26.6)
2478
(23.0)
1116
(30.2)
1.0
1972
(20.8)
1888
(33.2)
1.0
1385
(17.7)
2401
(33.4)
1.0
1.0
2.9 (1.9 to 4.6)
0.05
1.3 (1.2 to 1.4)
54 (10.6)
1.7 (1.6 to 1.8) 36 (25.5)
1.0
2.7 (1.7 to 4.1) 0.03
2.1 (1.9 to 2.2)
3707
0.8 (0.7 to 0.9) 85 (12.9)
(25.2)
753 (31.2) 1.3 (1.2 to 1.5) 21 (25.6)
2342
1.2 (1.1 to 1.3) 32 (25.6)
(27.7)
0.5 (0.3 to 0.8) 0.04
2.3 (1.4 to 3.7) 0.03
2.7 (1.7 to 4.2) 0.0004
All statistically significant (Po 0.05) hazard ratios and interactions are highlighted in bold text; for the South Asian and Black groups, we present only those hazard ratios
that were significantly different to the hazard ratio in the White group, according to formal testing of the statistical interaction term.
When we conducted further analysis to see if the results were altered by fitting an interaction term ‘other’ ethnic group to the analysis in the combined group, we found
the results remained essentially unchanged.
a
HRs are adjusted for centre
Data are at least 84% complete for all variables except for marital status (South Asian 76% complete; Black 80% complete; White 71% complete), living arrangements
(South Asian 69%; Black 75%; White 64%) and problems with alcohol misuse being a precipitant to the self-harm (South Asian 57%; Black 51%; White 68%). Variables where
there was a small number of events within South Asian or Black groups (fewer than 10 individuals repeating) to perform a statistical test for interaction were excluded;
these related to the following problems precipitating self-harm: relationship problems with those other than partner, friends or family, bereavement, employment/study
problems, financial, housing or legal problems, problems due to drug misuse, physical health problems and abuse (physical, sexual or emotional).
b
with mental health as a direct precipitant to self-harm increased
the risk of repetition. Factors associated with increased risk of
repetition in the BME groups compared to the White group were
less varied when including multiple repetitions by the same
individual. However, this analytical approach, modelling selfharm repetition count data, accentuated the strength of
association with alcohol misuse in both BME groups, and also
not having a partner among South Asians.
This is a large cohort study using data collected from three
centres in England with comprehensively ascertained follow-up
data on repetition of self-harm and completed suicide. However,
we only collected data on self-harm attendances to hospital and did
438
J. Cooper et al. / Journal of Affective Disorders 148 (2013) 435–439
Table 2
Negative binomial regression models showing incidence rate ratios (IRRs) for self-harm repetition countsy following index self-harm episode, with significant interactions
between risk factor and ethnic group.
Variable
b
White, Black and SA
Count of
repeats
IRRa (95% CI)
South Asian
Count of
repeats
Total
10 553
Sociodemographic:
In employment or education
Unemployed
7507
3046
1.0
1.2 (1.0 to 1.5)
Married or partnered
No partner
1697
5428
1.0
40
1.4 (1.3 to 1.4) 131
Lives with family/friends
Lives alone or homeless
5428
2164
1.0
1.6 (1.5 to 1.8)
1715
2081
2858
1.4 (1.2 to 1.6) 54
1.4 (1.2 to 1.6)
0.6 (0.5 to 0.6)
2080
0.9 (0.8 to 1.1)
Factors precipitating selfharm:
Problem alcohol use
Mental health symptoms
Relationship problems with
partner
Relationship problems with
family/friends
Black
IRRa (95% CI)
p value for
interaction term
(SA vs. White)
232
Psychiatric treatment and previous self-harm:
No self-harm in the past year 5224
1.0
Self-harm in the past year
3379
2.3 (2.1 to 2.5) 76
No previous self-harm
Self-harm4 1 year ago
5878
2745
1.0
1.4 (1.2 to 1.7)
Not in current treatment
In current treatment
4227
5114
1.0
2.1 (1.8 to 2.4)
No previous treatment
Any previous treatment
2696
6383
1.0
2.5 (2.4 to 2.6)
Circumstances of the act:
Self-poisoning involved
Self-cutting/stabbing involved
Alcohol used in the act
8729
2054
5631
0.7 (0.7 to 0.7)
1.5 (1.5 to 1.6)
1.3 (1.2 to 1.3) 95
IRRa (95% CI)
p value for
interaction term
(Black vs. White)
25
3.7 (2.2 to 6.3)
o 0.001
66
2.3 (1.6 to 3.4)
o 0.001
Count of
repeats
127
2.4 (1.7 to 3.5)
0.01
8.8 (5.9 to 13.2) o0.001
3.3 (2.5 to 4.5)
0.01
4.3 (3.2 to 5.8)
o0.001
All statistically significant (Po 0.05) IRRs and interactions are highlighted in bold text; for the South Asian and Black groups, we present only those IRRs that were
significantly different to the IRR in the White group, according to formal testing of the statistical interaction term.
a
IRRs are adjusted for centre
Data are at least 84% complete for all variables except for marital status (South Asian 76% complete; Black 80% complete; White 71% complete) and problems with
alcohol misuse being a precipitant to the self-harm (South Asian 57%; Black 51%; White 68%).
y
A maximum of ten repeat episodes were included for each individual.
b
not record repeat episodes that did not come to the medical
attention of the participating hospitals. Whilst the overall patient
sample was large, the number of suicides in the BME groups was
small and the number of repeaters within ethnic minority groups
somewhat limited the study’s statistical power. It is probable that
the smaller number of individuals in the Black group in particular
resulted in fewer statistically significant interactions for this group.
We included fatal self-harm repetitions (suicides) and non-fatal
repetitions as one outcome group (‘repeaters’), because we considered it important not to exclude the suicides from the analysis. The
suicides constituted only 5% of the whole outcome group and
therefore did not exert any undue influence on the findings overall.
Ethnicity is notoriously difficult to ascribe and, although we used
broad categories which may have reduced error, this approach could
have concealed differences between ethnic groups within the
categories we used. The relatively large number of individuals for
whom ethnicity was not recorded may have introduced a degree of
selection bias.
Although risk factors for repetition of self-harm appear to be
broadly similar across ethnic groups, the specific heterogeneity
findings need further explanation. It is unclear why mental health
symptoms should be a stronger risk factor in Black people; it may
be related to small numbers but could be due to differentiation in
their perception of and amenability to treatment for depression
(Cooper et al., 2010; Brown et al., 2011). Evidence of psychiatric
morbidity was a strong risk factor in the South Asian group as was
not having a partner, which may be an indication of mental
instability and/or lack of social support. Perhaps the stigma
attached to accessing psychiatric treatment (Bowl, 2007) means
only the acutely ill re-present to hospital. The effect of alcohol
misuse was stronger in South Asians, perhaps due to lower
baseline prevalence for this risk factor. In the South Asian
population people who drink may be more marginalised than
those who drink in the White population. Problematic alcohol use
within a small minority of South Asian groups and as a feature
within the South Asian self-harm population has been demonstrated previously (Bhogal et al., 2006). Alcohol appears to be a
problem in both BME groups for those who present with multiple
episodes of self-harm. This adds weight to our recommendation
that in BME groups, clinicians should ensure adequate recognition
and treatment of mental illness and alcohol misuse at assessment.
Our previous study (Cooper et al., 2010) showed that BME groups
were least likely to receive a psychosocial assessment following
presentation to hospital with self-harm and a specialist assessment
has been shown to be protective in reducing repetition (Kapur et al.,
2008; Bergen et al., 2010b). The findings from this study reinforce
J. Cooper et al. / Journal of Affective Disorders 148 (2013) 435–439
governmental recommendations (National Institute for Health and
Clinical Excellence (NICE), 2004) that all presentations should
receive a psychosocial assessment and this should include BME
groups. Barriers to accessing services for BME groups include the
stigma of mental illness within families and the community,
concerns about the treatment on offer (Gater et al., 2010) and
attributing depression solely to social factors (Brown et al., 2011). It
has been suggested that pathways to care for ethnic minority groups
should therefore be ethnic specific and rely more heavily on nonstatutory sector services (Moffat et al., 2009). Information to build
on guidance for service development is limited although new
evidence on interventions in specific minority ethnic groups is
emerging (Gater et al., 2010). The challenge for statutory services
in facilitating effective interventions for self-harm in BME groups is
to build collaborations between sectors sympathetic to cultural
beliefs and assist referral routes between these services.
Role of funding source
This is an independent report on research funded by the Department of Health
and the Policy Research Programme in the Department of Health. The views
expressed are not necessarily those of the Department.
Conflict of interest
KH is a National Institute for Health Research Senior Investigator. NK is the
Chair of the Guideline Development Group for the National Institute for Health
and Clinical Excellence (NICE) guideline on the longer-term management of selfharm. No other authors have any conflict of interest.
Acknowledgements
The authors thank their respective research teams, clinical and administrative
staff in Oxford, Manchester and Derby for assistance with self-harm data
collection.
References
Bergen, H., Hawton, K., Waters, K., Cooper, J., Kapur, N., 2010a. Epidemiology and
trends in non-fatal self-harm in three centres in England: 2000–2007. The
British Journal of Psychiatry 197, 493–498.
Bergen, H., Hawton, K., Waters, K., Cooper, J., Kapur, N., 2010b. Psychosocial assessment
and repetition of self-harm: the significance of single and multiple repeat episode
analyses. Journal of Affective Disorders 127, 257–265, http://dx.doi.org/10.1016/
j.jad.2010.05.001.
Bhogal, K., Baldwin, D., Hartland, L., Nair, R., 2006. Brief communication: differences between ethnic groups in demographic and clinical features of patients
439
admitted and assessed after deliberate self-harm: a retrospective case-note
study. The International Journal of Social Psychiatry 52, 483–486.
Bhugra, D., Arya, P., 2005. Ethnic density, cultural congruity and mental illness in
migrants. International Review of Psychiatry 17 (2) 133–137.
Bhui, K., McKenzie, K., Rasul, F., 2007. Rates, risk factors, and methods of self harm
among minority ethnic groups in the UK: a systematic review. BMC Public
Health 19 (7), 336.
Bowl, R., 2007. The need for change in UK mental health services: South Asian
service users’ views. Ethnicity and Health 12, 1–19.
Brown, J., Casey, S., Bishop, A., Prytys, M., Whitttinger, N., Weinman, J., 2011. How
Black African and White British women perceive depression and help-seeking:
a pilot vignette study. The International Journal of Social Psychiatry 57,
362–374.
Cooper, J., Husain, N., Webb, R., Waheed, W., Kapur, N., Guthrie, E., Appleby, L.,
2006. Self-harm in the UK: differences between South Asians and Whites in
rates, characteristics, provision of service and repetition. Social Psychiatry and
Psychiatric Epidemiology 41, 782–788.
Cooper, J., Murphy, E., Webb, R., Hawton, K., Bergen, H., Waters, K., Kapur, N., 2010.
Ethnic differences in self-harm, rates, characteristics and service provision: a
cohort study comparing three English cities. The British Journal of Psychiatry
197, 212–218.
Gardner, W., Mulvey, E.P., Shaw, E.C., 1995. Regression analyses of counts and
rates: Poisson, overdispersed Poisson, and negative binomial models. Psychological Bulletin 118, 392–404.
Gater, R., Waheed, W., Husain, N., Tomenson, B., Aseem, S., Creed, F., 2010. Social
intervention for British Pakistani women with depression: randomised controlled trial. The British Journal of Psychiatry 197, 227–233.
Kapur, N., Cooper, J., King-Hele, S., Webb, R., Lawlor, M., Rodway, C., Appleby, L.,
2006. The repetition of suicidal behavior: a multicenter cohort study. The
Journal of Clinical Psychiatry 67, 1599–1609.
Kapur, N., Murphy, E., Cooper, J., Bergen, H., Hawton, K., Simkin, S., Casey, D.,
Horrocks, J., Lilley, R., Noble, R., Owens, D., 2008. Psychosocial assessment
following self-harm: Results from the Multi-Centre Monitoring of Self-Harm
Project. Journal of Affective Disorders 106, 285–293.
Moffat, J., Sass, B., Mckenzie, K., Kamaldeep, B., 2009. Improving pathways into
mental health care for black and ethnic minority groups: A systematic review
of the grey literature. International Review of Psychiatry 21, 439–449.
Murphy, E., Kapur, N., Webb, R., Purandare, N., Hawton, K., Bergen, H., Waters, K.,
Cooper, J., 2012. Multicentre cohort study of older adults who have harmed
themselves: risk factors for repetition and suicide. The British Journal of
Psychiatry 200, 399–404.
National Institute for Health and Clinical Excellence (NICE), 2004. The short-term
physical and psychological management and secondary prevention of selfharm in primary and secondary care. Clinical Guideline 16.
Nazroo, J., Jackson, J., Karlsen, S., Torres, M., 2007. The Black diaspora and health
inequalities in the US and England: does where you go and how you get there
make a difference? Sociology of Health and Illness 29 (6), 811–830.
Statacorp, 2007. Stata Statistical Software: Release 10.: Stata Corporation, College
Station.
Steeg, S., Kapur, N., Webb, R., Applegate, E., Stewart, S.L., Hawton, K., Bergen, H.,
Waters, K., Cooper, J., 2012. The development of a population-level clinical
screening tool for self-harm repetition and suicide: the ReACT Self-Harm Rule.
Psychological Medicine 42, 2383–2394..